Hepatitis C virus (HCV) is considered to be the major etiological agent of non-A non-B (NANB) hepatitis, chronic liver disease, and hepatocellular carcinoma (HCC) around the world, with an estimated human seroprevalence of 1% globally. [Alter et al., 1994, Gastroenterol. Clin. North Am. 23:437-455; Behrens et al., 1996, EMBO J. 15:12-22]. Four million individuals may be infected in the United States. The viral infection accounts for greater than 90% of transfusion-associated hepatitis in the U.S. and it is the predominant form of hepatitis in adults over 40 years of age. Almost all of the infections result in chronic hepatitis and nearly 20% of those infected develop liver cirrhosis.
The virus particle has not been identified due to the lack of an efficient ex vivo replication system and the extremely low amount of HCV particles in infected liver tissues or blood. However, molecular cloning of the viral genome has been accomplished by isolating the messenger RNA (mRNA) from the serum of infected chimpanzees and preparing cDNA using recombinant methodologies. [Grakoui A. et al., 1993, J. Virol. 67: 1385-1395]. It is now known that HCV contains a positive strand RNA genome comprising approximately 9400 nucleotides, organization of which is similar to that of flaviviruses and pestiviruses. The genome of HCV, a (+)-stranded RNA molecule of .about.9.4 kb, encodes a single large polyprotein of about 3000 amino acids which undergoes proteolysis to form mature viral proteins in infected cells.
Cell-free translation of the viral polyprotein and cell culture expression studies have established that the HCV polyprotein is processed by cellular and viral proteases to produce the putative structural and nonstructural (NS) proteins. At least ten mature viral proteins are produced from the polyprotein by specific proteolysis. The order and nomenclature of the cleavage products are as follows: NH.sub.2 --C-E1-E2-p7-NS2-NS4A-NS3-NS4B-NS5A-NS5B-COOH (FIG. 1) [Grakoui et al., 1993, J. Virol. 67:1385-95; Hijikata et al., 1991, PNAS 88:5547-51; Lin et al., 1994, J. Virol. 68:5063-73]. The three amino-terminal putative structural proteins, C (capsid), E1, and E2 (two envelope glycoproteins), are believed to be cleaved by a host signal peptidase of the endoplasmic reticulum (ER). The host enzyme is also responsible for generating the amino terminus of NS2. The proteolytic processing of the nonstructural proteins are carried out by the viral proteases: NS2-3 and NS3, contained within the viral polyprotein. The NS2-3 protease catalyzes the cleavage between NS2 and NS3. It is a metalloprotease and requires both NS2 and the protease domain of NS3.
The NS3 protease catalyzes the rest of the cleavages in the nonstructural part of the polyprotein. The NS3 protein contains 631 amino acid residues and is comprised of two enzymatic activities: the protease domain contained within amino acid residues 1-181 and a helicase ATPase domain contained within the rest of the protein Kim et al., 1995, Biochem Biophys Res. Comm., 215:160-166. It is not known if the 70 kD NS3 protein is cleaved further in infected cells to separate the protease domain from the helicase domain, although no cleavage has been observed in cell culture expression studies.
The NS3 protease is a member of the serine class of enzymes. It uses a His, Asp, Ser catalytic triad. Mutation of the Ser residue abolishes cleavage of NS3/4A, NS4A/4B, NS4B/5A, and NS5A/5B substrates. The cleavage between NS3 and NS4A is intramolecular, whereas the cleavages at the NS 4A/4B, 4B/5A, 5A/5B sites occur in trans.
Experiments using transient expression of various forms of HCV NS polyproteins in mammalian cells have established that the NS3 serine protease is necessary but not sufficient for efficient processing of all of these cleavages. Like the flaviviruses, the HCV NS3 protease also requires a cofactor to catalyze some of these cleavage reactions. Efficient proteolytic processing at NS3/4A, NS4A/4B, NS4B/5A, and NS5A/5B sites within the non-structural domain of hepatitis C virus requires a heterodimeric complex of the NS3 serine protease and the NS4A protein. [Bartenschlager et al. 1995, J. Virol. 67:3835-3844; Failla et al., 1994, J. Virol. 68:3753-3760]. A 13-amino acid synthetic NS4A peptide, corresponding to the central hydrophobic domain of NS4A protein, spanning residues 21-33 has been shown to be sufficient for activation of NS3 protease [Butkiewicz et al., 1996, Virology, 225: 328-338]. A smaller domain (amino acid residues 22-30) of NS4A has been shown to be sufficient for activation of the protease [Lin et al., 1995, J. Virol 69:4377-80].
The recently published three dimensional structure of the NS3 protease [Kim et al, 1996, Cell 87:343-355; Love et al, 1996, Cell 87:331-342] revealed that the N-terminal 37 residues of NS3 adopt a .beta. (residues 6-9)-.alpha. (residues 14-22)-.beta. (residues 33-37) structure upon binding of a synthetic peptide corresponding to the central hydrophobic domain spanning residues 21-32 of NS4A protein.
Production of an active NS3.sub.1-181 -NS4A peptide complex at present involves two steps. First, the NS3 catalytic domain (amino acid residues 1-181) is produced as a recombinant protein in E. coli. Next, a 13-19 residue NS4A peptide spanning the central hydrophobic domain of the full-length NS4A protein is added to form a non-covalent complex [Kim et al., 1996, Cell 87:343-355]. This complex, although more active than the protease alone, is approximately 8-10 fold less active than the full-length NS3.sub.1-631 -NS4A.sub.1-54 form of the protease as judged by its proteolytic activity toward a synthetic substrate based on the native NS5A-NS5B amino acid sequence. [Urbani et al., 1997, J. Biol. Chem., 272(14):9204-09; Steinkuhler et al., 1996, J. Virol. 70(10):6694-6700]. Moreover, NS4A peptide has been shown to have a very low affinity (10 .mu.M) for NS3 in solution [Bianchi et al,. 1997, Biochemistry 36: 7890-7897], requiring addition of NS4A peptide in the high micromolar range to insure a 1:1 stoichiometric complex with NS3 protease. The limited solubility of this peptide in aqueous buffer due to its hydrophobic nature makes working with this peptide at these concentrations difficult.
Because the HCV NS3 protease cleaves the non-structural HCV proteins necessary for HCV replication, the NS3 protease can be a target for the development of therapeutic agents against the HCV virus. The gene encoding the HCV NS3 protein has been cloned as disclosed in U.S. Pat. No. 5,371,017. To date, however, the protease has not been produced in a covalent complex with the NS4A cofactor in a soluble, active and stable form. Such a complex would be useful as a target in a high throughput screen to discover therapeutic agents. A stable, active HCV protease is also required for determination of modes of binding of inhibitors by NMR, for structural determination by NMR spectroscopy, for crystallography, and for virtually all biophysical and biochemical studies interested in the activated form of the enzyme.